Piperazine-Based CCR5 Antagonists as HIV-1 Inhibitors. IV. Discovery of 1-[(4,6-Dimethyl-5-pyrimidinyl) carbonyl]-4-[4-{2-methoxy-1(R)-4-(trifluoromethyl)-phenyl}ethyl-3(S) -methyl-1-piperazinyl]-4-methylpiperidine (Sch-417690/Sch-D), a Potent, Highly Sel

Article abstract of DOI:10.1021/jm0304515

The nature and the size of the benzylic substituent are shown to be the key to controlling receptor selectivity (CCR5 vs M1, M2) and potency in the title compounds. Optimization of the lead benzylic methyl compound 3 led to the methoxymethyl analogue 30,

Full text of DOI:10.1021/jm0304515

J . Med. Chem. 2004, 47, 2405-2408
2405
Letters
P ip er a zin e-Ba sed CCR5 An ta gon ists a s
HIV-1 In h ibitor s. IV. Discover y of
1-[(4,6-Dim eth yl-5-p yr im id in yl)ca r bon yl]-
4-[4-{2-m eth oxy-1(R)-4-(tr iflu or om eth yl)-
p h en yl}eth yl-3(S)-m eth yl-1-p ip er a zin yl]-
4-m eth ylp ip er id in e (Sch -417690/Sch -D),
a P oten t, High ly Selective, a n d Or a lly
Bioa va ila ble CCR5 An ta gon ist
F igu r e 1. Piperidine and piperazine-based CCR5 Antago-
nists.
J ayaram R. Tagat,* Stuart W. McCombie,
Dennis Nazareno, Marc A. Labroli, Yushi Xiao,
Ruo W. Steensma, J ulie M. Strizki,
Bahige M. Baroudy, Kathleen Cox, J ean Lachowicz,
Geoffrey Varty, and Robert Watkins
their immune systems are apparently not compromised,
supporting the blockade of CCR5 as a mechanism for
inhibiting HIV.5
We have previously reported on the development of
two related series of compounds as CCR5 antagonists
(Figure 1). From the piperidino-piperidine series emerged
compound 1 (Sch-351125/Sch-C), which is currently in
clinical trials as the first orally active agent to block
HIV entry.6 The corresponding compound (2) in the
piperazino-piperidine series7 satisfied many of the initial
criteria but showed acute CV and CNS side effects in
ancillary pharmacology. This compound possessed mod-
est affinity for the muscarinic receptors. In addition to
brain, M1 and M3 muscarinic receptors are also preva-
lent in the GI tract, and M2 receptors are expressed in
heart and lung tissues. We surmised that the high oral
blood levels that we deemed necessary for an anti-
infective agent such as 2 were not compatible with even
modest affinities for other receptors. Additionally, com-
pound 2 exhibited four rotational isomers (rotamers)
arising from hindered rotations about the two bonds
forming the unsymmetrical tertiary piperidino-amide
moiety. One pair of these rotamers can be rendered
degenerate by making the amide symmetrical, with the
4,6-dimethylpyrimidine carboxamide (3) being optimal.8
Although the CV profile of the pyrimidine carboxamide
3 was much better than that of the N-oxide 2 in rats,
some GI effects were still noted at higher doses of 3.
Hence, we decided to further refine the template rep-
resented by 3 with the goal of minimizing or eliminating
its affinity for the muscarinic receptors and perhaps
improving its CCR5 affinity. The following is an account
of the key aspects of our back-up program which
resulted in the discovery of the titled compound 30, our
second generation orally active CCR5 antagonist, which
is currently in clinical trials as an HIV entry inhibitor.9
The piperidino-piperidine 1 and the piperazino-pip-
eridine 2 and 3 have similar scaffolds and the SAR is
very similar for the two ends of the molecules. However,
there are distinguishing features to the two structures
in the middle which contributes to the differences in
their activity. Hence we focused our attention on the
central portion of the piperazino-piperidine structure
with the view of modulating receptor selectivity. Replac-
ing the 2(S)-methyl group on the piperazine ring with
an ethyl or propyl group resulted in substantial loss of
binding at CCR5, and so we turned our attention to the
Schering-Plough Research Institute, K-15-2B-2800,
2015 Galloping Hill Road, Kenilworth, New J ersey 07033
Received September 10, 2003
Abstr a ct: The nature and the size of the benzylic substituent
are shown to be the key to controlling receptor selectivity
(CCR5 vs M1, M2) and potency in the title compounds.
Optimization of the lead benzylic methyl compound 3 led to
the methoxymethyl analogue 30, which had excellent receptor
selectivity and oral bioavailability in rats and monkeys.
Compound 30 (Sch-417690/Sch-D), a potent inhibitor of HIV-1
entry into target cells, is currently in clinical trials.
Despite the advances in our knowledge of the patho-
genesis of the human immunodeficiency virus type 1
(HIV-1) and in therapeutic intervention, the AIDS
pandemic continues to be the globe’s leading public
health issue. Although the inhibitors of reverse tran-
scriptase and viral protease enzymes have contained the
disease progression in HIV patients, the development
of resistance to these drugs highlights the need for new,
mechanistically novel therapies.1 About seven years ago,
the chemokine receptors CCR5 and CXCR4 were identi-
fied as essential coreceptors for the attachment of HIV-1
to the surface of CD4+ macrophages and T-cells.2 The
coreceptor engagement leads to a conformational change
in the gp 41 glycoprotein on the viral envelope, revealing
a peptide which fuses the HIV envelope to the cell
membrane, thus aiding viral entry into the cell. The
chemokine receptors CCR5 and CXCR4 which control
this key event in the HIV entry process belong to the
extensively studied super family of G-protein-coupled
receptors (GPCRs), making them attractive targets for
drug design.3 Studies of HIV tropism indicated that the
R5 viral strains that utilize CCR5 are responsible for
the initial establishment and development of the dis-
ease, which prompted several research groups to design
small molecule ligands to block the HIV-CCR5 interac-
tion.4 Individuals lacking functional CCR5 receptors
(∆32-homozygotes) are resistant to HIV infection while
* To whom correspondence should be addressed. Phone: (908)-740-
3461; fax: (908)-740-7441; e-mail: jayaram.tagat@spcorp.com.
10.1021/jm0304515 CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/09/2004

2406 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 10
Ta ble 1. Larger Benzylic Substituents Improve Receptor Selectivitya
Letters
muscarinic Ki (nM)
M2
456
rat PK (po)
c
R in 22
CH3
CH3CH2
CH3CH2CH2
PhCH2
CCR5 Ki (nM)
M1
575
3900
6075
M3
716
9089
>5300
>5300
HIV entry (nM)b
AUC0-6h
3
2.8
5.2
1.1
1.4
0.39
0.18
0.35
0.13
6210
7310
4600
2030
12
20
21
4750
4760
>5700
>3900
a
b
For descriptions of antiviral, muscarinic and rat PK assays, please see refs 7 and 8. IC50 values for inhibiting the entry of replication
defective HIV-1 (YU-2) in to U-80 cells. c Oral blood levels of compound in rats (10 mg/kg, n)2). Area under curve ) ng/mL × h.
Sch em e 1. The SN2 Displacement Route
Sch em e 2. The Cyanoamine Alkylation Route
Ta ble 2. Larger Benzylic Substituents Enhance Potency vs
HIV-Isolatesa
IC90 (nM) values for the inhibition
of HIV-1 clinical isolates in PBMC
R in 22
J rFL ADA-M 301657 J V1083 RU 570
1
3
oxime-piperidine 19
25
19
21
0.03
0.04
2.8
43
13
16
0.09
0.6
1.8
10
31
18
0.27
0.25
4.9
400
>1000
1000
CH3
40
19
12 CH3CH2
20 CH3CH2CH2
21 PhCH2
0.01
0.19
3.3
0.43
4
10
30 CH3OCH2
benzylic substituent, leaving the 2(S)-methyl piperazino-
piperidine intact. Two different scalable synthetic routes
were developed to facilitate the systematic variation of
the benzylic substituent.
a
For a description of antiviral assays, please see refs 6 and 7.
(OAc)3H] resulted in the formation of the benzylic propyl
derivative 16, as a 1:1 mixture of diastereomers. Flash
chromatography served to isolate the desired (S,S)
diastereomer which was parlayed to the target 20. The
benzyl group was similarly introduced by treating the
cyanoamine anion with benzyl bromide, decyanation,
and further elaboration to 21. Propyl and larger iodides
failed to alkylate the hindered cyanoamine anion.
Targets bearing ether side chains at the benzylic
position were obtained by the SN2 displacement route,
as the alkoxy alkyl halides were not reactive enough to
alkylate the cyanoamine. For example (Scheme 1),
R-methoxyacetophenone 25 was made efficiently by the
addition of aryllithium 23 to the Weinreb amide 24.12
CBS reduction of the R-alkoxyacetophenone, activation
to a mesylate, and reaction with monoprotected chiral
piperazine 7 proceeded in high overall yields. Comple-
tion of the sequence as before secured the target 30.
Even with a one carbon homologation to the ethyl
derivative 12, we observed (Table 1) an order of mag-
nitude improvement in receptor selectivity (CCR5 vs
M1-M3) that was further improved with larger groups
Initially, we approached the synthesis of the benzylic
ethyl analogue by adapting our published SN2 displace-
ment route (Scheme 1).7 Thus, the commercially avail-
able propiophenone 4 was reduced by the CBS protocol10
to furnish benzylic alcohol 5 enriched in the R-enanti-
omer. Activation to the mesylate 6 and reaction with
the chiral piperazine 7 led to the desired (S,S) diaster-
eomeric product 8, which was processed to the final
target 12.8 Extension of the SN2 displacement route to
larger benzylic substituents was inefficient.
Our second synthetic strategy (Scheme 2) exploited
the alkylation of a cyanoamine, followed by decyanation
to introduce larger benzylic substituents.11 Commercial
aryl aldehydes were condensed with the chiral pipera-
zine 7 under modified Strecker conditions6,7 to obtain
the cyanoamines 14. Treatment of 14 with NaHMDS
under scrupulously anaerobic conditions and quenching
with allyl bromide provided the R-allyl cyanoamine 15
in good yield. Saturation of the allyl group (H2/Pd-C)
and reductive removal of the nitrile [MgBr2‚OEt2/NaB-

Letters
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 10 2407
Ta ble 3. Analogues of the n-Propyl Compound
muscarinic
Ki (nM)
rat PKc
CCR5
entryb
IC50 (nM)
HIV-PBMCa
IC90 (nM)
R in 22
Ki (nM)
M1
M2
AUC0-6h
20
30
31
32
CH3CH2CH2
CH3OCH2
Cyc-Pr-CH2
CF3CH2CH2
1.1
2.5
1.6
ND
6075
>10000
1894
4760
>10000
3512
0.35
0.46
0.8
0.03-0.45
1-10
0.01-4
0.4-20
4600
6900
5900
4600
>3700
>6700
1.44
a
HIV isolates included: J rCSF, ASM57, Bal, J V1083, and RU570. b,c Please refer to footnotes b, c in Table 1 for details.
Ta ble 4. Pharmacokinetic Profiles of Compounds 20 and 30
such as propyl (20) and benzyl (21). However, we had
to balance the increase in receptor selectivity against
the observed oral blood levels of these benzylic ana-
logues in our high throughput pharmacokinetic (PK)
screen in rats.7 Compounds bearing lipophilic side
chains were metabolized readily by liver enzymes,
resulting in lower blood levels after oral administration.
Thus, the serum level for the propyl compound 20,
though acceptable, was much less than that of the ethyl
analogue 12, and the benzyl derivative 21 had the
lowest oral blood level among this group.
iv admin
oral admin
species
T1/2
BA
(%)c
a
b
a
(dose, mg/kg) AUC0-24h
(h) Cmax
AUC0-24h
20
30
rat (2)
monkey (2)
4550
2670
15.5
6.3
311
240
3270
1760
72
66
rat (10)
monkey (2)
21000
2240
8
3.5
2300
670
21000
1990
100
89
a
AUC (area under curve of concn vs time plot) ) ng/mL × h.
Cmax ) ng/mL. c BA (bioavailability) ) [AUC]po/[AUC]iv.
b
Another significant benefit that accrued from the
presence of larger benzylic substituents was enhanced
potency of inhibition of HIV-1 strains in peripheral blood
mononuclear cells (PBMCs). The IC90 (nM) values for
inhibiting genetically diverse HIV-1 isolates for the new
piperazino-piperidines were compared with the first
clinical candidate 1 (Table 2). In terms of their potency,
the methyl and ethyl derivatives were about equal to
1, but were not as broad spectrum in eliciting their
response. The effect of further carbon homologation from
ethyl to propyl was very dramatic, and the n-propyl
compound 20 was the most potent compound in this
panel. The benzyl analogue 21 was almost as potent as
20, but it had the lowest oral blood levels in rat among
this group (Table 1).
Complete profiling of the very potent propyl com-
pound 20 showed that this compound was orally well
absorbed in rat and monkey (Table 4). It did not inhibit
or induce liver enzymes, and was tolerated without
significant GI and CV/CNS side effects up to a dose of
10 mg/kg in rats. On the negative side, compound 20
was >99% protein bound in human plasma (compound
1 is only 84% protein bound), and during a chronic
dosing study in rats, lack of body weight gain relative
to control was observed at 10 and 30 mg/kg doses.
Further study showed that the propyl side chain was
oxidized to carboxylic acid which forms an acyl glucu-
ronide (observed in rat bile), a potentially toxic metabo-
lite.13
F igu r e 2. Potencies of compounds 1 and 30 vs HIV isolates.
compensated by improvements in the oral blood level
(Table 3) and protein binding (84%) for the latter. The
cyclopropylmethyl analogue 31 had the best potency in
this triad vs a small panel of isolates, but its receptor
selectivity (CCR5 vs M1 and M2) was diminished and
it was >99% protein bound. In a head-to-head chronic
dosing study in rats, there were no significant adverse
side effects with 30 whereas some lack of weight gain
and mild enzyme induction were noted with 31. Com-
pared to 30, the trifluoropropyl compound 32 showed
lower oral blood levels, lacked the broad spectrum
potency against diverse HIV isolates, and was 99%
protein bound.
Direct comparison (Figure 2) of compound 30 with our
first generation CCR5 antagonist 1 against a geographi-
cally diverse panel of HIV isolates (n ) 52) showed that
30 is an order of magnitude more potent than 1, with a
mean IC50 ) 0.45 nM and IC90 ) 4 nM.
Further work confirmed that the methoxymethyl
compound 30 had high oral bioavailability in rodents
and primates (Table 4), in addition to very good potency
and receptor selectivity. It has good CNS penetration
(brain/plasma ratio ) 0.86 for 30 and 0.17 for 1). The
major route of metabolism (rats) for 30 is through the
liver, where O-demethylation of the side chain followed
by glucuronidation was observed.
While we were developing compound 30 in the back-
up program, QT prolongation was observed with 1
(doses g 400 mg/day) in the electrocardiogram of
healthy human volunteers. This is thought to arise from
the interaction of the putative drug with the hERG
potassium ion channel in heart tissue.14 In the voltage
clamp assay,15 the piperazino-piperidine 30 had weaker
To block the oxidation of the terminal methyl group
in the propyl side chain, we prepared three related
derivatives. The methoxymethyl (30), cyclopropylmethyl
(31), and the 3,3,3-trifluoropropyl (32) analogues were
prepared via the SN2 displacement route. Comparison
of the data for these three compounds (Table 3) showed
that 30 had the best overall profile: Introduction of an
oxygen atom in the propyl side chain, in the form of the
methoxymethyl (MOM) derivative (30), reduced its
lipophilicity and essentially eliminated affinity for the
muscarinic receptors. The MOM compound 30 was more
potent than the first generation piperidino-piperidine
1, but less potent than the n-propyl derivative 20 (Table
2). However, the modest drop in potency in going from
the n-propyl (20) to the MOM compound (30) was

2408 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 10
Letters
(2) Berger, E. A.; Murphy, P. M.; Farber, J . M. Chemokine receptors
as HIV-1 co-receptors: roles in viral entry, tropism and disease.
Annu. Rev. Immunol. 1999, 17, 657-700.
(3) Hunt, S. W. III; LaRosa, G. J . Chemokine receptors as HIV co-
receptors: targets for therapeutic intervention in AIDS. Annu.
Rep. Med. Chem. 1998, 33, 263-272.
(4) Kazmierski, W.; Bifulco, N.; Yang, H.; Boone, L.; DeAnda, F.;
Watson, C.; Kenakin, T. Recent progress in discovery of small-
molecule CCR5 chemokine receptor ligands as HIV-1 inhibitors.
Bioorg. Med. Chem. 2003, 11, 2663-2676.
(5) Liu, R.; Paxton, W.; Choe, S.; Ceradini, D.; Martin, S.; Horuk,
R.; MacDonald, M.; Stuhlmann, H.; Koup, R.; Landau, N. R.
Homozygous defect in HIV-1 coreceptor accounts for resistance
of some multiply exposed individuals to HIV-1 infection. Cell
1996, 86, 367-377.
(6) Palani, A.; Shapiro, S.; Clader, J .; Greenlee, W.; Cox, K.; Strizki,
J .; Endres, M.; Baroudy, B. Discovery of 4-[(Z)-(4-Bromophenyl-
(ethoxyimino)methyl]-1′-[(2,4-dimethyl-3-pyridinyl)carbonyl]-4′-
methyl-1,4′-bipiperidine N-Oxide (SCH 351125): An Orally
Bioavailable Human CCR5 Antagonist for the Treatment of HIV
Infection. J . Med. Chem. 2001, 44, 3339-3342.
(7) Tagat, J . R.; Steensma, R.; McCombie, S.; Nazareno, D.; Lin,
S.-I.; Neustadt, B.; Cox, K.; Xu, S.; Wojcik, L.; Murray, M.;
Vantuno, N.; Baroudy, B.; Strizki, J . Piperazine-Based CCR5
Antagonsits as HIV-1 Inhibitors. II: Discovery of 1-[(2,4-
Dimethyl-3-pyridinyl)carbonyl]-4-methyl-4-[3(S)-4-methyl-4-[1(S)-
[4-(trifluoromethyl)phenyl}ethyl}-1-piperazinyl}piperidine N1-
Oxide (Sch-350634), an Orally Bioavailable, Potent CCR5
Antagonist. J . Med. Chem. 2001, 44, 3343-3346.
(8) McCombie, S.; Tagat, J . R.; Vice, S.; Lin, S.-I.; Steensma, R.;
Palani, A.; Neustadt, B.; Baroudy, B.; Strizki, J .; Endres, M.;
Cox, K.; Dan, N.; Chou, C.-C. Piperazine-Based CCR5 Antago-
nists as HIV-1 Inhibitors. III: Synthesis, Antiviral and Phar-
macokinetic Profiles of Symmetrical Heteroaryl Carboxamides.
Bioorg. Med. Chem. Lett. 2003, 13, 567-571.
(9) Tagat, J .; McCombie, S.; Nazareno, D.; Steensma, R.; Labroli,
M.; Baroudy, B.; Strizki, J .; Cox, K. Discovery of potent, selective
and orally active CCR5 antagonists for the potential treatment
of HIV infection. Abstracts of Papers; 225th ACS National
Meeting, New Orleans, LA, March 23-27, 2003; American
Chemical Society: Washington, D. C., 2003; MEDI 15.
(10) Corey, E. J .; Bakshi, R. K.; Shibata, S. Highly enantioselective
borane reductions of ketones catalyzed by chiral oxaborolidines.
J . Am. Chem. Soc. 1987, 109, 5551-5553.
hERG activity (IC50 ) 5.8 µM) than 1 (IC50 ) 1.1 µM).
In an ancillary study in cyanomolgus monkeys dosed
with compound 30 and monitored with cardiac telem-
etry, no cardiovascular effects (including QT prolonga-
tion) were observed up to a dose of 40 mg/kg. Upon
chronic dosing of 30 in rats, there was neither inhibition
nor induction of liver enzymes. Additionally, no acute
CNS or GI effects were noted in an ancillary study with
30 (10 mg/kg oral dose in rats). Thus our premise of
designing piperazino-piperidine leads with high receptor
selectivity (CCR5 vs M1, M2) as the basis for improved
side effect profile had fulfilled its promise.
In summary, during our back up program, we dis-
covered that the size of the benzylic substituent is the
key to enhanced potency and receptor selectivity in the
piperazino-piperidine series. These efforts initially led
to the n-propyl analogue 20 with excellent broad spec-
trum potency. However, the formation of an acylglucu-
ronide metabolite from this compound in vivo, and
undesirable side effects during chronic dosing, prompted
us to introduce a heteroatom (O) in the propyl side chain
to alter its metabolic fate. The resulting methoxymethyl
derivative 30 had the best overall profile among all the
compounds in the second phase of our program: It was
extremely selective for the CCR5 receptor with almost
no affinity for muscarinic receptors or other GPCRs. It
inhibited a broad spectrum of HIV isolates with a mean
IC90 of 4 nM, an order of magnitude more potent than
the first generation compound (1). The MOM compound
(30) had excellent oral bioavailability in rats and
monkeys, reduced affinity for the hERG K+ channel, and
no significant CNS or GI side effects at an oral dose of
10 mg/kg in rats. Compound 30 is currently in clinical
trials.
(11) (a) Ahlbrecht, H.; Raab, W.; Vonderheid, C. Dehydrocyanation
of R-aminonitriles: A versatile and convenient enamine and
dienamine synthesis. Synthesis 1979, 127-129. (b) Bunnelle, W.
H.; Shevlin, C. G. On the reduction of R-aminonitriles with
sodium. Tetrahedron Lett. 1989, 30, 4203-4206.
(12) Nahm, S.; Weinreb, S. W. N-methoxy-N-methyl amides as
effective acylating agents. Tetrahedron Lett. 1981, 22, 3815-
3818.
(13) Shipkova, M.; Armstrong, V. W. Oellerich, M.; Wieland, E. Acyl
glucuronide drug metabolites: toxicological and analytical im-
plications. Therap. Drug Monit. 2003, 25, 1-16.
(14) Pearlstein, R.; Vaz, R.; Rampe, D. Understanding the structure-
activity relationship of the Human Ether-a-go-go-Related Gene
cardiac K+ channel. A model for bad behavior. J . Med. Chem.
2003, 46, 1-6.
Ack n ow led gm en t. We thank Nicole Wagner, Lisa
Wojcik, and Serena Xu for antiviral data, Vilma Ruperto
for M1-M3 binding data, Dr. J esse Wong for the large
scale synthesis of key intermediates, and Dr. Steve
Sorota for the hERG data. We are grateful to Drs. J ohn
Piwinski and Greg Reyes for their strong support of this
program.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and spectral data for the preparation and character-
ization of compounds 12, 20, 30-32 are available free of charge
at http://pubs.acs.org.
(15) Fermini, B.; Fossa, A. The impact of drug-induced QT interval
prolongation on drug discovery and development. Nat. Rev. Drug
Discovery 2003, 2, 439-447.
Refer en ces
(1) Pomerantz, R. J . Primary HIV-1 resistance. J . Am. Med. Assoc.
1999, 282 (12), 1177-1178.
J M0304515